Antisense inhibition of gene expression relies primarily on the simple rules of Watson-Crick base pairing of nucleic acids. A synthetic small single-stranded oligodeoxynucleotide (ODN) (13-25 mer) that is complementary to a specific gene, via hybridizing to corresponding mRNA, inhibits the translation of that gene into a protein. Targeting gene expression at the RNA level gives cells another level of regulatory control, allowing them to turn off protein production even if RNA is abundant. If the protein product of translation were important for cell growth and/or viability, antisense inhibition of gene expression could produce a lethal phenotype. Unmodified phosphodiester ODNs are not suitable therapeutic agents because they are too readily digested by nucleases. To resolve this problem, several ODN analogs have been introduced, but phosphorothioate ODNs (PS-ODNs) have been extensively studied in various models and are now being tested in human clinical trials. Second-generation antisense ODNs that are superior to PS-ODNs have also been introduced. Hybridization of antisense ODNs to their target mRNAs can physically block the translation machinery or activate RNase H cleavage at the RNA-DNA duplex site. An extensive amount of literature points to the sequence-specific antisense mechanism of action at the single-gene level, but exploration of its effect on global gene expression in the cell have been scarce. We examined genomic effects of antisense inhibition of protein kinase A RIa expression in tumor cells using cDNA microarrays. Using in vivo tumor models of PC3M human prostate carcinoma grown in nude mice, the specificity of antisense effects on gene expression signatures was critically assessed using three oligonucleotides that differed in sequence or chemical modification: an immunostimulatory phosphorothioate oligonucleotide directed against human RIa, a second-generation immunoinhibitory RNA-DNA mixed-backbone ODN, and a non-immunostimulatory phosphorothioate ODN targeted to mouse RIa (this ODN cross-hybridizes with human RIa). Antisense treatment was found to downregulate one cluster, or signature, of genes involved in proliferation and upregulate another involved in differentiation. These expression signatures were quiescent and unaltered in the livers of antisense-treated animals, indicating that distinct cAMP signaling pathways regulate growth for normal and cancer cells. Thus,cDNA microarray analysis of the RIa antisense-induced expression profile exhibited the up- and downregulation of clusters of coordinately expressed genes that define the molecular portrait of a reverted tumor cell phenotype. This global view broadens the horizons of antisense technology; it advances the promise of antisense beyond a single target gene to the whole cell and the whole organism. Along with recent rapid advances in oligonucleotide technologies-including new chemical and biological understanding of more sophisticated nucleic acid drugs, oligonucleotide-based gene silencing offers not only an exquisitely specific genetic tool for exploring basic science but an exciting possibility for treating and preventing cancer and other diseases. Further adoption of this technology will facilitate development of nucleic acid medicines with higher target specificity and minimized side effects. There are two types of cyclic AMP (cAMP)-dependent protein kinase (PKA), type I (PKA-I) and type II (PKA-II), which share a common catalytic (C) subunit but contain distinct regulatory (R) subunits, RI versus RII, respectively. Evidence suggests that increased expression of PKA-I and its regulatory subunit (RIa) correlates with tumorigenesis and tumor growth. We investigated the effect of sequence-specific inhibition of RIa gene expression at the initial phase of 7,12-dimethylbenz(a)anthracene (DMBA)-induced mammary carcinogenesis. Antisense mixed-backbone ODN targeted against PKA RIa was administered (0.1 mg/day/rat, i.p.) on one day before DMBA intubation and during the first 9 days post-DMBA intubation to determine the effect on the early phase of carcinogenesis. Antisense RIa, in a sequence-specific manner, inhibited the tumor production. At 90 days after DMBA intubation, untreated controls and RIa-antisense-treated rats exhibited an average mean number of tumors per rat of 4.2 and 1.8, respectively, and 90% of control and 45% of antisense-treated animals had tumors. The antisense also delayed the first tumor appearance. An increase in RIa and PKA-I levels in the mammary gland and liver preceded DMBA-induced tumor production, and antisense down-regulation of RIa restored normal levels of PKA-I and PKA-II in these tissues. Antisense RIa in the liver induced the phase II enzymes, glutathione S-transferase and quinone oxidoreductase, c-fos protein, and activator protein 1 (AP-1)- and cAMP response element (CRE)-directed transcription. In the mammary glands, antisense RIa promoted DNA repair processes. In contrast, the CRE transcription-factor decoy could not mimic these effects of antisense RIa. The results demonstrate that RIa antisense produces dual anticarcinogenic effects: (a) increasing DMBA detoxification in the liver by enhancing phase II enzyme activities, CRE-binding-protein phosphorylation and CRE- and Ap-1-directed transcription; and (b) activating DNA repair processes in the mammary gland by down-regulating PKA-I.